Image Similarity Based on Intensity Using Mutual Information

In this paper, image similarity is finding based on intensity. Mutual information measures the amount of information that one image information contains about image information based on intensity. There are three ways to find mutual information using joint entropy, conditional entropy, and relative entropy. Entropy is find uncertainty of image. If images are more similar to other image, mutual information value is high and vice versa. If images are dissimilar or independent then mutual information is zero. When two images are exact same then mutual information value is maximum (same as entropy of an image).